Competition Dynamics Shape Algorithmic Phases of In-Context Learning

• URL: http://arxiv.org/abs/2412.01003v3
• Date: Sat, 28 Dec 2024 21:01:43 GMT
• Title: Competition Dynamics Shape Algorithmic Phases of In-Context Learning
• Authors: Core Francisco Park, Ekdeep Singh Lubana, Itamar Pres, Hidenori Tanaka,
• Abstract summary: In-Context Learning (ICL) has significantly expanded the general-purpose nature of large language models. We propose a synthetic sequence modeling task that involves learning to simulate a finite mixture of Markov chains. We show we can explain a model's behavior by decomposing it into four broad algorithms that combine a fuzzy retrieval vs. inference approach with either unigram or bigram statistics.
• Score: 10.974593590868533
• License:
• Abstract: In-Context Learning (ICL) has significantly expanded the general-purpose nature of large language models, allowing them to adapt to novel tasks using merely the inputted context. This has motivated a series of papers that analyze tractable synthetic domains and postulate precise mechanisms that may underlie ICL. However, the use of relatively distinct setups that often lack a sequence modeling nature to them makes it unclear how general the reported insights from such studies are. Motivated by this, we propose a synthetic sequence modeling task that involves learning to simulate a finite mixture of Markov chains. As we show, models trained on this task reproduce most well-known results on ICL, hence offering a unified setting for studying the concept. Building on this setup, we demonstrate we can explain a model's behavior by decomposing it into four broad algorithms that combine a fuzzy retrieval vs. inference approach with either unigram or bigram statistics of the context. These algorithms engage in a competition dynamics to dominate model behavior, with the precise experimental conditions dictating which algorithm ends up superseding others: e.g., we find merely varying context size or amount of training yields (at times sharp) transitions between which algorithm dictates the model behavior, revealing a mechanism that explains the transient nature of ICL. In this sense, we argue ICL is best thought of as a mixture of different algorithms, each with its own peculiarities, instead of a monolithic capability. This also implies that making general claims about ICL that hold universally across all settings may be infeasible.

Related papers

• Unified Generative and Discriminative Training for Multi-modal Large Language Models [88.84491005030316]
  Generative training has enabled Vision-Language Models (VLMs) to tackle various complex tasks. Discriminative training, exemplified by models like CLIP, excels in zero-shot image-text classification and retrieval. This paper proposes a unified approach that integrates the strengths of both paradigms.
  arXiv  Detail & Related papers  (2024-11-01T01:51:31Z)
• Dynamic Post-Hoc Neural Ensemblers [55.15643209328513]
  In this study, we explore employing neural networks as ensemble methods. Motivated by the risk of learning low-diversity ensembles, we propose regularizing the model by randomly dropping base model predictions. We demonstrate this approach lower bounds the diversity within the ensemble, reducing overfitting and improving generalization capabilities.
  arXiv  Detail & Related papers  (2024-10-06T15:25:39Z)
• LoRA-Ensemble: Efficient Uncertainty Modelling for Self-attention Networks [52.46420522934253]
  We introduce LoRA-Ensemble, a parameter-efficient deep ensemble method for self-attention networks. By employing a single pre-trained self-attention network with weights shared across all members, we train member-specific low-rank matrices for the attention projections. Our method exhibits superior calibration compared to explicit ensembles and achieves similar or better accuracy across various prediction tasks and datasets.
  arXiv  Detail & Related papers  (2024-05-23T11:10:32Z)
• Transformers are uninterpretable with myopic methods: a case study with bounded Dyck grammars [36.780346257061495]
  Interpretability methods aim to understand the algorithm implemented by a trained model. We take a critical view of methods that exclusively focus on individual parts of the model.
  arXiv  Detail & Related papers  (2023-12-03T15:34:46Z)
• Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection [88.23337313766353]
  This work first provides a comprehensive statistical theory for transformers to perform ICL. We show that transformers can implement a broad class of standard machine learning algorithms in context. A emphsingle transformer can adaptively select different base ICL algorithms.
  arXiv  Detail & Related papers  (2023-06-07T17:59:31Z)
• What and How does In-Context Learning Learn? Bayesian Model Averaging, Parameterization, and Generalization [111.55277952086155]
  We study In-Context Learning (ICL) by addressing several open questions. We show that, without updating the neural network parameters, ICL implicitly implements the Bayesian model averaging algorithm. We prove that the error of pretrained model is bounded by a sum of an approximation error and a generalization error.
  arXiv  Detail & Related papers  (2023-05-30T21:23:47Z)
• Streaming Inference for Infinite Non-Stationary Clustering [9.84413545378636]
  Learning from a continuous stream of non-stationary data in an unsupervised manner is arguably one of the most common and most challenging settings facing intelligent agents. Here, we attack learning under all three conditions (unsupervised, streaming, non-stationary) in the context of clustering, also known as mixture modeling. We introduce a novel clustering algorithm that endows mixture models with the ability to create new clusters online, as demanded by the data.
  arXiv  Detail & Related papers  (2022-05-02T21:05:18Z)
• Structured Reordering for Modeling Latent Alignments in Sequence Transduction [86.94309120789396]
  We present an efficient dynamic programming algorithm performing exact marginal inference of separable permutations. The resulting seq2seq model exhibits better systematic generalization than standard models on synthetic problems and NLP tasks.
  arXiv  Detail & Related papers  (2021-06-06T21:53:54Z)
• The Role of Isomorphism Classes in Multi-Relational Datasets [6.419762264544509]
  We show that isomorphism leakage overestimates performance in multi-relational inference. We propose isomorphism-aware synthetic benchmarks for model evaluation. We also demonstrate that isomorphism classes can be utilised through a simple prioritisation scheme.
  arXiv  Detail & Related papers  (2020-09-30T12:15:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.